LIDAR image of swamp in brazil
LIDAR image of swamp in Brazil

Reforestation Goes High-Tech With Pachama

When Argentinian entrepreneur Diego Saez Gil landed himself between projects in 2017, he went to the Madre de Dios area of the Peruvian Amazon – and what he observed there ignited an idea that would turn into an extraordinary tech start-up aimed at making reforestation a more practical tool for climate mitigation.

According to figures from Global Forest Watch, deforestation totaled 72.65 million acres in 2017 – equal to an area more than half the surface area of France. This is in spite of a number of efforts to use carbon offsets to conserve forests and promote reforestation. International ventures like REDD+ –diminutive for reduced emissions from deforestation and forest degradation– target to reward state and non-state landowners for keeping their forests rather than cutting them down.

Regrettably, says Saez Gil, there are two main challenges with projects like this: timely, precise measuring of the carbon being kept and getting the takings of carbon offset payments to communities on the ground. “First, in order to measure how much carbon is being sequestered in a certain forest area, traditionally, you have had to go these distant areas and gauge trees with measuring tapes and infer from there.”

pachamas forest plans

It takes months and fieldwork and costs $100,00 to $300,000 to do these measurements using conventional methods

“These same techniques are used around the world, even in California, one of the world’s tech capitals,” he said, adding that varying measurements and the fact that these calculations are usually only done once every five years can lead to an absence of trust in forest carbon projects. Saez Gil says his company Pachama, established in 2018, is the remedy to the problem: The platform considers data collected by satellite imaging and other remote-sensing technologies, then uses artificial intelligence to generate useful figures.

“In experiments in California, our technology had a 1.5% error rate in contrast to the US Forestry Service statistics,” he said, adding that the procedure takes a few days compared to numerous months using traditional methods.

One of the technologies motivating this method is light detection and ranging (LiDAR) attached to aircraft or drones. LiDAR tech is reaching to the extent where it can enter tropical forest canopies and give an exact, 3-D snapshot of the forest.

reforestation technology
Satellite imagery is also beneficial in revealing the density of the forest and surrounding areas.

“We are working on an assignment in Brazil, its 70,000 hectares, you can’t observe that physically,” Saez Gil said. That remote-sensing data is then fed into AI technologies which crunch the data into a trustworthy figure of how much carbon is being seized. “We use deep learning, and we use neural networks,” he further added.

According to a study published in the journal Science, the restoration of large areas of forested land across the world could be the international community’s best gamble in assisting to alleviate climate change by seizing carbon from the atmosphere.

But the second big challenge in executing that, Saez Gil says, is that there are frequently many middlemen between the clients in the developed world seeking to offset their carbon emissions and the local forest managers who are keepers of the areas where the forest carbon is being deposited.

YOU MAY LIKE: The 5 Environmental Stories That Could Outline 2020

Saez Gil says he wishes his company’s technology and marketplace platform will allow local groups, farmers, and a much broader range of people to reforest their land and to be better rewarded for doing so.

“For instance, if you are a farmer in Colombia who is keeping an area of forest, before, you had to go through this expensive and prolonged process to get their carbon certificates and find a purchaser, “he said,

“Now, that same farmer could have their forest verified with our technology and then be able to move on the platform and find a purchaser without having to go through an agent.”